This proposal is a revision of the competing continuation reviewed by the Study Section in October 1995. The study section raised a number of questions, and these are addressed both in the introduction and throughout the text of the proposal. The investigators also report on additional progress made in the original grant. The setting of this study is the Cardiovascular Health Study (CHS), a cohort study of 5888 older adults designed to assess risk factors for stroke and coronary heart disease. In CHS, cardiovascular risk factors and medical conditions are well characterized, and participants are followed closely for events. The investigators point out that the CHS provides an excellent opportunity to monitor use of and changes in cardiovascular medications among community-dwelling older adults. Over the last several decades, the results of randomized controlled clinical trials have helped to define optimal care for patients with cardiovascular disease or with many of the risk factors for cardiovascular disease. Many reviews are available, and guidelines have been formulated for the care of patients with specific conditions such as atrial fibrillation or asthma or with risk factors such as elevated levels of cholesterol. But information about the implementation of guidelines or the effect of clinical trials on clinical practice patterns is often lacking; and when such data are available, they tend to be too old to say much about "current" practice. The aim of this project is to describe the use of various cardiovascular medications among older adults. In some instances, it is temporal trends that are of primary interest. The number of important questions is large. These include the use of: (1) ACE inhibitors in patients with CHF; (2) lipid-lowering drugs in coronary patients; (3) beta-blockers in post-MI patients; (4) ACE inhibitors in diabetic patients; (5) aspirin in patients with TIA, angina or MI; (6) calcium-channel blockers in patients with coronary disease; (7) aspirin and warfarin in patients with atrial fibrillation; (8) estrogens and progestins in women; (9) anti-arrhythmic agents; (10) ACE inhibitors in renal disease; (11) anti-hypertensive agents; and (12) asthma medications. For each group of agents or for each condition, the investigators can look at drug use cross-sectionally as well as drug use over time. The ability to examine dose in CHS also enables them to identify undertreatment, which is common in older adults. Having characterized the subjects who are not treated or who are undertreated, they can also identify the predictors of non-treatment and undertreatment. In the current paper on cholesterol-lowering agents, for instance, in which hypertension increased the likelihood of initiation of lipid-lowering drug therapy among drug-therapy eligible subjects, none of the other components of the NCEP guidelines--gender, diabetes, smoking, obesity, family history of premature CHD, and total number of risk factors--was associated with the initiation of lipid-lowering drug therapy. The investigators point out that knowledge about the parts of guidelines that are ignored or underused is essential to developing interventions to improve compliance with those guidelines. They further state that tracking progress toward the implementation of guidelines in a timely fashion is an important method of continually drawing attention to the scientific evidence underlying those guidelines.